Editor's Note: This is the 2nd blog post in a series on Cloud, Kubernetes, and Dataware. The first post in the series, "Rent or Buy: Should You Go to Cloud (or Not)?"
How useful is the cloud?
The balance between advantages or disadvantages of cloud services depends in large part not on any particular cloud vendor but on you: What is the nature of your business and what are the particular requirements? Whether or not the cloud is a good choice — and when, and what strategies — will be different for different types of business or at different points in an organization's growth and maturity. Only when you have a good fit between your own situation and your cloud strategy are you likely to find the advantages you seek.
The same tradeoffs apply across each business, but their relative impact is different. Thinking about a range of businesses and the difference in advantages for cloud — especially at the extremes — gives you a sense of how the fit between business and cloud strategy works.
With that in mind, take a quick look through these examples to see if you find yourself in any of them:
- Start up: Cloud can mean less hassle and investment time and less initial cost if there are not already established infrastructure and IT resources for a new company. A cloud-first and initially a cloud-only policy make mean faster time for development and faster time to business value. For start-ups, depending on the level of experience for the team, there's a learning curve in the business itself. You may be able to delay or possibly entirely avoid the effort of setting up management for your own cluster(s). There's also an issue of flexibility: you may not yet know enough about the future of your business to know what you're going to want, so on-demand services from the cloud may be a good fit. Or maybe you're going to have a sharp ramp-up time as may happen with an online business that suddenly becomes widely known, and you don't want to buy too much hardware ahead of when it's needed. Cloud may be a good option for you.
- Small business: Like the start-up, maybe you don't want the effort or the cost of setting up your own IT team to run on-premises clusters or you want to avoid the cost of physical space for servers. The actual cost is one issue; a distraction from key business goals is another. Maybe cloud costs a bit more but keeps your attention where it needs to be. Cloud all or in part may be a good option. But if this is you, do your homework: do research the actual costs of cloud as you envision using it in the present and near future and weigh that against on-premises costs. Also, what restrictions or regulations may your company face? Will cloud meet your needs?
- Business with lots of money but a special focus: Maybe you can afford the costs of cloud services even if they become larger than expected, but there are other considerations. Suppose your company has special requirements such as deep learning systems that need particular hardware (think GPUs). This is not always sufficiently available in the cloud. Or you may face regulatory issues. Hybrid architecture may be the best strategy.
- Medium sized business: For businesses in the middle of the range, there may be the most variability as to whether or not cloud will be advantageous or when in a business's growth cloud is really a good choice. This is the case where your particulars must drive the decision based on good fundamental understanding; there are not good general predictors.
- Big business with lots of experience and lots of funds: Going to the cloud may not actually cost less, but that's not the only reason to go to the cloud. You may like what cloud offers in terms of the freedom to experiment and the flexibility, especially if you follow the recommendations in Parts II and III of this book to ensure you maintain current portability and future flexibility in your choices.
- Huge business: Just as Google and Amazon can run a cloud less expensively than each individual group could do, a huge business with lots of money and lots of expertise might as well just do this themselves. You could have the economies of scale and the flexibility to experiment yet maintain control by building your own large system. But remember: there are only a few companies who fit into this category. Even some of them have gone to the cloud or to multiple clouds although they have the resources to do as they like.
Did you find your business in the preceding list? Whether you have already decided on using public cloud resources or do not plan to do so at present, the cloud is a trend you most likely need to understand, either for your own use or because your customers have data and applications in the cloud.
The blog posts in this series are intended to help you do that. Watch for new blog posts in the series as they are published. Other resources on the topic are listed here.
For more information on these topics try these free resources:
Blog Post: "Rent or Buy: Should You Go to Cloud (or Not)?" by Ellen Friedman, 1st in the Cloud, Kubernetes and Dataware series
Whiteboard Walkthrough video with Ted Dunning "Big Data in the Cloud"
eBook: Kubernetes for Machine Learning, Deep Learning and AI by Sam Charrington
eBook: AI and Analytics in Production by Ted Dunning & Ellen Friedman
Whitepaper: "Realizing the Full Potential of Your Cloud Investment"
Try MapR: MapR Live! to try MapR free via a live web-based learning environment or launch MapR in your favorite cloud provider
This blog post was published June 11, 2019.